www.nature.com/scientificreports OPEN Assessing the susceptibility of schools to food events in Iran Saleh Yousef1, Hamid Reza Pourghasemi2*, Sayed Naeim Emami1, Omid Rahmati3, Shahla Tavangar4, Soheila Pouyan2, John P. Tiefenbacher5, Shahbaz Shamsoddini1 & Mohammad Nekoeimehr1 Catastrophic foods cause deaths, injuries, and property damages in communities around the world. The losses can be worse among those who are more vulnerable to exposure and this can be enhanced by communities’ vulnerabilities. People in undeveloped and developing countries, like Iran, are more vulnerable and may be more exposed to food hazards. In this study we investigate the vulnerabilities of 1622 schools to food hazard in Chaharmahal and Bakhtiari Province, Iran. We used four machine learning models to produce food susceptibility maps. The analytic hierarchy process method was enhanced with distance from schools to create a school-focused food-risk map. The results indicate that 492 rural schools and 147 urban schools are in very high-risk locations. Furthermore, 54% of rural students and 8% of urban students study schools in locations of very high food risk. The situation should be examined very closely and mitigating actions are urgently needed. Floods are among the destructive natural hazards. Tese extreme events can be generated by a number of natural processes or from human activities and catastrophes, including heavy precipitation events, melting snowpack, modifed drainage networks, failures of dams, and manipulation of drainage features. Based on recorded data, foods have caused US $700 billion globally and about 7 million deaths since 1900 1. Floods are about 30% of hazardous events2,3. During last few decades, urbanization and increasing in populations have greatly increased exposure of people and properties to foods 4–6. Some studies indicate that food frequency and severity may increase as a consequence of global warming and changing climates7,8. Floods could be managed and mitigated by sof (nature-based and/or non-structural) and hard (engineered and structural) actions and decisions. Te hard actions include dams, diversions, and check dams. Sof actions include land use planning, river restoration, selective siting of buildings, food prediction modeling, alarm systems, improving public awareness of food hazards, and education9,10. Floods infuence soil erosion, enhance natural habitats, support ecological processes, and are important to many aspects of human life. In communities, children are the most vulnerable to the consequences of food exposure 11,12. Schools are set- tings that concentrate children and need special attention with regard to extreme natural events. Over the last few decades, the frequency of foods has been increasing and loss of life and property has accordingly increased13–15. So, it is important to assess susceptibility of schools to food events to reduce damages and prevent loss of lives. For this purpose, a food susceptibility and hazard map can be prepared using various techniques or algorithms including statistical and machine learning. Machine learning (ML) algorithms like logistic regressions 16–18, random forests19,20, support vector machines21–24, decision trees 25,26, artifcial neural networks 27,28, boosted regression trees 29–31, multivariate adap- tive regression splines 29,32, and model-driven architectures 16,33 have been tested for hazard analysis and mapping in literature. Te ML approach has been used to evaluate the risk and susceptibility of communities exposed to a number of extreme and hazardous conditions: landslides 34–36, wildfres37,38, gully erosion processes39–41, land subsidence42,43, earthquakes4,13,44, dust storms45, and foods6,7,46. Flood-hazard vulnerability has been examined by a number of scholars. Ochola et al. 47 studied the susceptibility of schools to foods in the Nyando River basin in Kenya. Tey analyzed the conditions of 130 schools in the western part of that country and found that 40% were vulnerable to foods. Karmakar et al.48 conducted a risk-susceptibility analysis of foods in southwestern Ontario, Canada. Tey evaluated four types of vulnerability—physical, economic, infrastructural, and social—using a 1Soil Conservation and Watershed Management Research Department, Chaharmahal and Bakhtiari Agricultural and Natural Resources Research and Education Center, AREEO, Shahrekord, Iran. 2Department of Natural Resources and Environmental Engineering, College of Agriculture, Shiraz University, Shiraz, Iran. 3Soil Conservation and Watershed Management Research Department, Kurdistan Agricultural and Natural Resources Research and Education Center, AREEO, Sanandaj, Iran. 4Department of Watershed Management Engineering, College of Natural Resources, Tarbiat Modares University, Tehran, Iran. 5Department of Geography, Texas State University, San Marcos, TX, USA. *email: [email protected] Scientifc Reports | (2020) 10:18114 | https://doi.org/10.1038/s41598-020-75291-3 1 Vol.:(0123456789) www.nature.com/scientificreports/ Figure 1. Flowchart of the methodology in present study. geographic information system (GIS). Balica et al.49 examined food susceptibility using parametric and physical models and concluded that parametric modeling has limited accuracy, but provides a simplifed view of social indicators of vulnerability. Nabegu50 studied the vulnerabilities of households to fooding in Kano, Nigeria. Tey found that houses in the most vulnerable zone were destroyed and 17 people lost their lives during food events. Eini et al.51 investigated urban food susceptibility using ML techniques in Kermanshah, Iran. Tey prepared food maps using two ML models—maximum entropy and genetic algorithm—and found that maximum entropy yielded a more accurate food-susceptibility model. Tey also determined that infrastructural characteristics had the greatest infuence on food susceptibility. Tascón-González et al.52 studied social food-vulnerability in Ponferrada, Spain using analytic hierarchy process (AHP) and found that 34,941 residents were impacted by foods from a dam break, and that 77% of them sufered heavy damages. Few have attempted to examine the susceptibility of school locations to foods. A risk assessment of schools in developing countries is very important but has not yet been conducted. Tis study is the frst to investigate the exposure of both urban and rural schools to food hazards. It has been conducted for the mountainous province of Chaharmahal and Bakhtiari, Iran. Te goal is to identify the locations most in need of mitigation to reduce damages and prevent loss of lives. Four ML models were tested and compared for the tasks of mapping food hazard and assessing schools’ exposures. Materials and methods Study area. Chaharmahal and Bakhtiari Province is in southwestern Iran in a region dominated by the Zagros Mountains. Having an average elevation of 2153 m above sea level and a range of elevations from 778 to 4203 m, the province is the highest in Iran. Te province covers 16,421 km2 and its population is approximately 947,000. Due to the topographical and climatic conditions of the region, foods occur annually throughout the province. Methodology. Tere are fve steps to this research: (1) collection and compilation of spatial data; (2) deter- mination of the infuence of the independent efective factors on food probability; (3) production of food risk maps using four ML algorithms; (4) validation and evaluation of the food risk maps, and (5) determination of the susceptibility of schools to foods in Chaharmahal and Bakhtiari Province (Fig. 1). Scientifc Reports | (2020) 10:18114 | https://doi.org/10.1038/s41598-020-75291-3 2 Vol:.(1234567890) www.nature.com/scientificreports/ Figure 2. Locations of the foods that occurred between 1977 and 2019 in Chaharmahal and Bakhtiari Province and visual examples of several events. Collection and compilation of spatial data. To accurately determine food patterns and frequencies in a region, an accurate and well-distributed sample of food occurrence must be compiled. Tree hundred and forty-six foods that occurred in the province were recorded over a 42-year period (1977–2019) by Iran’s Ministry of Energy. Te locations of the foods were identifed and geo-located using a global position system (GPS) device during extensive feld surveys. Tese points were mapped (Fig. 2). Te sample was randomly divided into a modeling set containing 70 percent of the locations and a validation set containing 30% of the sample. As food occurrence is determined by an interaction of natural and human processes, based on previous studies 15,53–55 12 of the most important efective factors were identifed for use in modeling as input variables. Tey included elevation, slope, aspect, plan curvature, lithology, drainage density, annual rainfall, topographic wetness index (TWI), normalized diference vegetation index (NDVI), land use type, distance from nearest river, and distance from nearest road. Te data were derived from 1:25,000 topographic maps, 1:100,000 geological maps, and OLI Landsat images (from 2018). Te 12 data layers were created in ArcGIS 10.4.2 and ENVI 5.3 sofware. To ensure that the 12 input factors were truly independent of each other (not highly correlated with each other), a multicol- linearity test was applied. Te Pearson correlation tests showed no signifcant correlation between the factors, ensuring a more accurate food risk map (Fig. 3). Determination of the infuence of input factors on food probability. Some topographic factors can interact to